PLFS data: What the numbers hide

The third periodic labour force survey (PLFS) data for the year 2019-20 have been released recently. The biggest surprise is the increase in the labour force and work force participation rates compared to the last two rounds of annual averages.

In fact, if we keep in view the quinquennial surveys (NSSO: 2004-05 or 2011-12), the first PLFS witnessed a decline in the workforce participation rate (WFPR), particularly for women, and a sharp increase in the unemployment rate of both males and females. This was interpreted as an outcome of demonetisation which presumably reduced the scale of economic activities.

However, the next round of PLFS data for the year 2018-19 also did not show any marked improvement on the job front except some marginal decline in the unemployment rate.

Following the pandemic and the lockdown the quarterly data published for the urban sector showed a massive decline in the WFPR and a huge increase in the unemployment rate in the April-June quarter of the year 2019-20. Despite this, the annual average figures of WFPR rose sharply and the average unemployment rate declined by one percentage point from 5.8 per cent in the previous year to 4.8 per cent in 2019-20. This is where the suspicion grows stronger.

We agree that the WFPR was affected adversely more in the urban areas, and the unemployment rate rose phenomenally subsequent to the lockdown. As the migrants left for their rural homes, many of them must have started working at least on subsidiary basis. However, the rural job opportunities could not have neutralised the urban adversities in spite of the MNREGA.

One may argue that the sharp improvements in the labour market indicators during the first three quarters of the year 2019-20 may have bypassed the deterioration observed in the fourth quarter and thus, the average annual figures are strikingly positive. But what significant changes had occurred during the first three quarters of the year 2019-20 to make such a positive change in the annual average?

The fineprint

Let us look at some of the indicators carefully. The labour force participation rate (LFPR) in the rural areas recorded an all-time high in 2019-20 among the males.

Even compared to the NSS rounds, the rural male LFPR increased to 56.3 which is higher than the 2004-05 rate. Among the females, after witnessing a major decline in the year 2017-18 and 2018-19 the rate went up to 24.7 per cent, quite similar to the 2011-12 figure.

In the urban areas the male specific rate is again around 1 percentage point higher than the average figure observed since 2004-05 and among the females it is again an all-time high in 2019-20. The WFPR both in the rural and urban rates rose compared to the previous PLFS and almost approached the figures recorded in the earlier NSS round (2011-12). Among the urban females it is even marginally higher than the 2004-05 figure which was at the highest till 2018-19.

The unemployment rates between 2017-18 and 2018-19 were not much different except among the urban females. But the 2019-20 figures almost uniformly show a one percentage point decline among the rural and urban males as well as females. Had the pandemic not occurred, the participation rates would have been much higher and the unemployment rates, much lower!

The first three quarter estimates were so much improved that the large-scale decline in livelihood opportunities in the fourth quarter, especially in the urban areas, could do nothing to deteriorate the annual estimates (Figure 1). Rural opportunities seem to have compensated for the urban livelihood loss, which is too optimistic to hold. Hence, the logical conclusion that strikes the mind is as follows:

Either we say that the PLFS (2017-18 and 2018-19) are not comparable with the NSS estimates and 2019-20 tries to overcome the shortfalls;

Or we say that the PLFS (2017-18 and 2018-19) and NSS figures are compatible and start disbelieving the 2019-20 results.

The latter seems more convincing because questioning the compatibility of NSS and PLFS (2017-18 and 2018-19) is not all that easy. In fact, it becomes more swaying as we look at the figures based on principal status concept.

The participation rates improved and the unemployment rate declined between 2018-19 and 2019-20. All we get to learn from the theoretical standpoint that the participation rates do not increase on a year-to-year basis. It is only in the long run that the participation rates change.

Hence, it becomes difficult to establish the superiority of the government database over the Centre for Monitoring Indian Economy (CMIE) dataset which is presumably susceptible to sampling and many other problems.

Another line of cross-verification can be made before doubting the 2019-20 data.

The States’ factor

The participation rates at the all-India level are State averages. So, a regional decomposition which is underlying the aggregate estimates can be examined by envisaging the changes in the labour market indicators across the States and Union Territories.

Among the males, only some of the North-Eastern States such as Assam, Meghalaya and Manipur and from the south, Kerala and Puducherry reported a decline in the WFPR in 2019-20 in comparison to the 2018-19 estimates.

Among the females only Andhra, Delhi, Goa, Meghalaya, Chandigarh and Puducherry registered a decline; otherwise, the rest showed an increase. And the range of increase is really shattering.

Even if we ignore the figures for some of the UTs like Lakshadweep (more than 17 and 10 percentage point increase in the work participation rate of males and females respectively), several major States such as Gujarat, Jharkhand, Karnataka, Madhya Pradesh, Maharashtra, Odisha and Uttarakhand reported an unreconcilable increase in the work participation rate of females over just one year.

Even the male work participation rate increased to an extent of more than 2 percentage points in a number of major States. Disqualifying the theoretical underpinnings such massive increase in the annual work participation rates is definitely shocking.

Mitra is Professor at Institute of Economic Growth

COMMENT NOW